Machine Learning Approaches for Epidemiological Investigations of Food-Borne Disease … · 2019. 9. 23. · Food-Borne Disease Outbreaks Baiba Vilne1,2*, Irena Meistere¯ 1, Lelde

MINI REVIEWpublished: 06 August 2019

doi: 10.3389/fmicb.2019.01722

Frontiers in Microbiology | www.frontiersin.org 1 August 2019 | Volume 10 | Article 1722

Edited by:

Sophia Johler,

University of Zurich, Switzerland

Reviewed by:

Laura M. Carroll,

Cornell University, United States

Heather A. Carleton,

Centers for Disease Control

and Prevention (CDC), United States

*Correspondence:

Baiba Vilne

[email protected]

Specialty section:

This article was submitted to

Food Microbiology,

a section of the journal

Frontiers in Microbiology

Received: 07 March 2019

Accepted: 12 July 2019

Published: 06 August 2019

Citation:

Vilne B, Meistere I, Grantiņa-Ieviņa L

and Ķibilds J (2019) Machine Learning

Approaches for Epidemiological

Investigations of Food-Borne Disease

Outbreaks. Front. Microbiol. 10:1722.

doi: 10.3389/fmicb.2019.01722

Machine Learning Approaches forEpidemiological Investigations ofFood-Borne Disease OutbreaksBaiba Vilne 1,2*, Irēna Meistere 1, Lelde Grantiņa-Ieviņa 1 and Juris Ķibilds 1

1 Institute of Food Safety, Animal Health and Environment—“BIOR,” Riga, Latvia, 2 SIA net-OMICS, Riga, Latvia

Foodborne diseases (FBDs) are infections of the gastrointestinal tract caused by

foodborne pathogens (FBPs) such as bacteria [Salmonella, Listeria monocytogenes

and Shiga toxin-producing E. coli (STEC)] and several viruses, but also parasites and

some fungi. Artificial intelligence (AI) and its sub-discipline machine learning (ML) are

re-emerging and gaining an ever increasing popularity in the scientific community and

industry, and could lead to actionable knowledge in diverse ranges of sectors including

epidemiological investigations of FBD outbreaks and antimicrobial resistance (AMR). As

genotyping using whole-genome sequencing (WGS) is becoming more accessible and

affordable, it is increasingly used as a routine tool for the detection of pathogens, and

has the potential to differentiate between outbreak strains that are closely related, identify

virulence/resistance genes and provide improved understanding of transmission events

within hours to days. In most cases, the computational pipeline of WGS data analysis

can be divided into four (though, not necessarily consecutive) major steps: de novo

genome assembly, genome characterization, comparative genomics, and inference of

phylogeny or phylogenomics. In each step, ML could be used to increase the speed

and potentially the accuracy (provided increasing amounts of high-quality input data) of

identification of the source of ongoing outbreaks, leading to more efficient treatment and

prevention of additional cases. In this review, we explore whether ML or any other form

of AI algorithms have already been proposed for the respective tasks and compare those

with mechanistic model-based approaches.

Keywords: machine learning, food-borne disease, outbreaks, bacterial WGS, bioinformatics analysis pipeline

1. INTRODUCTION

Foodborne diseases (FBDs) are infections of the gastrointestinal tract caused by foodbornepathogens (FBPs) such as bacteria and several viruses, but also parasites and some fungi.Salmonella, Listeria monocytogenes and Shiga toxin-producing Escherichia coli (STEC) are someof the most important bacterial FBPs (Sekse et al., 2017), causing the most outbreaks and thelargest number of sporadic cases with severe illness or even fatal outcome (EFSA, 2015; Sekseet al., 2017). Salmonella infections affect people at all ages and the main food sources of infectiontypically include ready-to-eat foods, eggs, swine and poultry. L. monocytogenes infections mostlyaffect elderly people, as well as immunocompromised patients and pregnant women, and displayhigh mortality rates. Common food sources of L. monocytogenes include ready-to-eat foods suchas smoked fish and soft cheeses. STEC has been associated with severe complications, e.g., acutekidney failure, often affecting elderly and immunocompromised people, and also small children.

https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.org/journals/microbiology#editorial-boardhttps://www.frontiersin.org/journals/microbiology#editorial-boardhttps://www.frontiersin.org/journals/microbiology#editorial-boardhttps://www.frontiersin.org/journals/microbiology#editorial-boardhttps://doi.org/10.3389/fmicb.2019.01722http://crossmark.crossref.org/dialog/?doi=10.3389/fmicb.2019.01722&domain=pdf&date_stamp=2019-08-06https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articleshttps://creativecommons.org/licenses/by/4.0/mailto:[email protected]://doi.org/10.3389/fmicb.2019.01722https://www.frontiersin.org/articles/10.3389/fmicb.2019.01722/fullhttp://loop.frontiersin.org/people/542326/overviewhttp://loop.frontiersin.org/people/699803/overviewhttp://loop.frontiersin.org/people/751249/overview

Vilne et al. ML for Food-Borne Disease Outbreaks

The main food sources of STEC infections are bovine meat,followed by vegetables and juice (EFSA, 2015).

Whole-genome sequencing (WGS) is becoming moreaccessible and affordable as a routine approach for earlydetection of FBD outbreaks (Buultjens et al., 2017; Sekse et al.,2017). WGS captures the entire genome within hours to days andhas the potential to differentiate between outbreak strains that areclosely related, identify virulence/resistance genes and provideimproved understanding of transmission events (Quainoo et al.,2017; Andersen and Hoorfar, 2018). Moreover, third-generationsequencing technologies such as Oxford Nanopore (ONT)sequencing and PacBio Single Molecule, Real-Time (SMRT),which allow the generation of ultra-long (up to 300 kb) reads,are well suited to assemble reference genomes from outbreakstrains de novo, potentially contributing to more precisetaxonomic assignment, while offering increased detectionspeed and relatively decreasing costs, as, in comparison toIllumina short-read sequencing, both technologies are still threeand almost seven times more expensive, respectively (Brownet al., 2017; Sekse et al., 2017; Nicola De Maio, 2019). Severalproof-of-concept studies have demonstrated the superiorityof WGS over traditional typing methods for a range of highpriority food-borne pathogens, e.g., Salmonella enterica, Listeriamonocytogenes, Campylobacter species and STEC (Kanamoriet al., 2015; Quick et al., 2015; Moran-Gilad, 2017). Largeinitiatives have emerged to investigate the options of replacingconventional methods with WGS for outbreak investigations.Two such examples include the ENGAGE (Establishing NextGeneration sequencing Ability for Genomic analysis in Europe)(Hendriksen et al., 2018) and INNUENDO projects (Llarenaet al., 2018), focusing on the idevelopment of dedicated analyticalplatforms and standardized analysis pipelines, e.g., for E. coli anddifferent Salmonella spp. serotypes (Hendriksen et al., 2018).

In the era of Big Data, as the volume and complexity ofdata increases steadily, artificial intelligence (AI) and its sub-discipline machine learning (ML) are re-emerging and gainingan ever increasing popularity in the scientific communityand industry (Ching et al., 2018). While mechanistic model-based approaches aim at constructing simplified mathematicalformulations, i.e., hypothesis, of causal mechanisms by carefullyobservating, analyzing and trying to understand the complexityof the respective phenomenon (Baker et al., 2018), machinelearning (ML) algorithms use large-scale datasets to extractmeaningful patterns (i.e., “learn”) and use this “knowledge” tomake predictions on other data (Alkema et al., 2016). Moreover,ML can be done in a unsupervised manner by exploring anddetecting patterns within the data or in a supervised mannerby classifying, predicting and explaining (Tebani et al., 2016).Unsupervised ML techniques involve well-known and widelyused methods such as principal component analysis (PCA) andk-means clustering (Tebani et al., 2016). PCA is a dimensionalityreduction method, transforming a large set of variables into asmaller set, while preserving as much information as possible(Hotelling, 1933), whereas k-means clustering groups similardata points together in a fixed number (k) of clusters and tries todiscover their underlying patterns (Hartigan andWong, 1979). Inlife sciences, some frequently used supervised ML strategies have

been Random Forest (RF), Support Vector Machines (SVM),Naive Bayes (NB), and Artificial Neural Networks (Lai et al.,2016). RF alorithm randomly selects a subset from the trainingdata to construct an ensemble of decision tree predictors toaggregate the predictions, thus lowering the variance (Breiman,1996). SVM represent a pattern classification technique, whichis based on the idea of transforming the original data thatis not linearly separable to a higher dimensional space andfinding a hyperplane separating the data into classes (Boseret al., 1992). NB represents a probabilistic algorithm that usesthe probability theory and Bayes’ Theorem in conjunction withprior knowledge to calculate the probability of each feature tobelong to each of the classes and then outputs the class withthe highest probability (Devroye et al., 2013). Finally, ANNs aregraph computing models, which, at least to some extent, shouldmimic the functioning of the human brain, hence its computingunits are called neurons and are interconnected for passinginformation to each other. Moreover, networks of neurons areadditionally organized in layers. The first one is an input layer,receiving the training data. This is followed by several hiddenlayers. The last one is an output layer, which performs the actualprediction of the class (Kruse et al., 2016).

Global multi-disciplinary initiatives like One Health(OH) (http://www.onehealthinitiative.com/), aiming towardoptimizing the health of people, animals and the environment,would greatly profit from such approaches, as multiple complexchallenges need to be addressed, including the maintenance ofa safe food and water supply for a growing human population.Considering the current ease with which people and animalsor animal products can be transported around the globe,the forefront issues of OH are clearly related to spread ofemerging infectious diseases and antimicrobial resistance(AMR) (Gibbs, 2014). Especially, outbreaks caused by multi-drug-resistant bacteria are an urgent and growing globalpublic health threat (CDC, 2013; WHO, 2014). Effectivemanagement protocols must be in place, as quick identificationleads to faster and more precisely targeted treatment(Quainoo et al., 2017).

ML strategies have already been used for microbial diagnosticsin diverse contexts, including (i) taxonomic grouping ofmetagenomics data (Sedlar et al., 2017; Afify and Al-Masni,2018); (ii) classification of L. monocytogenes persistence in retaildelicatessen environments (Vangay et al., 2014); (iii) phenotypeprediction of bacterial strains based on presence/absence ofparticular genes (i.e., gene-trait matching) (Dutilh et al., 2013;Alkema et al., 2016; Farrell et al., 2018); (iv) to identify strainsthat demonstrate a higher probability to cause severe diseases(Wheeler et al., 2018); (v) to predict the host range of pathogens(Lupolova et al., 2017), e.g., identifying their signatures ofhost adaptation (Wheeler et al., 2018); and (vi) to predict theantimicrobial resistance potential of different E. coli strains (Herand Wu, 2018) or from different sources (Li et al., 2018).

The WGS data analysis pipeline can be generally dividedinto four major steps (Figure 1): de novo genome assembly,genome characterization, comparative genomics and inference ofphylogeny or phylogenomics (Quainoo et al., 2017). However,these steps are not necessarily consecutive, depending on the


http://www.onehealthinitiative.com/https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


FIGURE 1 | An overview of an example bacterial sequence data analysis workflow.

objectives of the study. ML could be used in any of theseanalyses to increase the speed and potentially accuracy (providedincreasing amounts of high-quality input data). In this review,we aim to explore whether ML algorithms have already beenproposed for the respective task and compare those algorithmswith mechanistic model-based approaches (see Table 1 foran overview). We mainly focus on single-genome short-read(Illumina) bacterial WGS; however in cases where, to the bestof our knowledge, no ML algorithms have been reported forthe respective task, we also briefly touch upon ML algorithmsdedicated to ultra-long read technologies, 16S metataxonomicsand shotgun metagenomics, as these approaches may find futureapplications in FBD outbreaks. Currently, the starting pointfor any FBD outbreak investigation involving strain typingis access to isolates, which may be difficult to obtain orare often even unavailable. Moreover, most food samples arecomplex, harboring composite microbial communities. In thisregard, metagenomic approaches would allow one to capturethe full spectrum of microbes in foods entirely without priorneed for culturing and isolation, allowing also the detectionof “viable but not cultivable," as well as non-viable microbes(Bergholz et al., 2014).

2. MACHINE LEARNING FOR DE NOVOMICROBIAL GENOME ASSEMBLY

Genome assembly tools are applied with the purpose ofassembling the sequencing reads into larger fragments (i.e.,

contigs), from which near-complete genomes can be furtherre-constructed. As the read lengths of the second generation(e.g., Illumina) technologies are short (i.e., 50–300 bp), de novoassembly without a reference genome remains a challengingtask (Zhu et al., 2014). However, de novo assembly is especiallyrelevant in FBD outbreak investigations, where the source strainmight be undetectable with conventional methods and thustaxonomically unclassified (Quainoo et al., 2017). Currently, themajority of the algorithms are based on the de Bruijn graphor overlap-layout strategies. The de Bruijn graph algorithmfirst splits up each read into smaller substrings, k-mers,which are further used to construct a graph, in which k-mers represent nodes; two nodes are connected with anedge if they overlap by k-1 nucleotides and follow eachother in the read. Thus, each contig is represented as apath within the graph (Zhu et al., 2014). The overlap-layout-based algorithms start by computing the overlaps amongall the reads, which are then used to perform the genomeassembly (Zhu et al., 2014). For short Illumina read-basedsingle genome WGS, the most popular assemblers includeVelvet (Zerbino and Birney, 2008), IDBA-UD (Peng et al.,2012), RAY (Boisvert et al., 2010), SPAdes (Bankevich et al.,2012), and SKESA (Souvorov et al., 2018), all of which employthe de Bruijn graph-based assembly strategy. The overlap-layout-based algorithms are mainly used for the assemblyof ultra-long reads: Minimap/miniasm (Li, 2016) and Canu(Koren et al., 2017).

For 16S metataxonomics data, interestingly, there is atool REAGO (REconstruct 16S ribosomal RNA Genes from


https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


TABLE 1 | An non-exhaustive list of the mechanistic model-based vs. ML tools for microbial genome analysis.

Category Tools

Mechanistic model-based Machine learning

DE NOVO GENOME ASSEMBLY

Velvet (Zerbino and Birney, 2008), IDBA-UD (Peng et al., 2012),

RAY (Boisvert et al., 2010), SPAdes (Bankevich et al., 2012),

SKESA (Souvorov et al., 2018) Minimap/miniasm (Li, 2016), Canu

(Koren et al., 2017), REAGO** (Yuan et al., 2015)

PERGA (Zhu et al., 2014), Minimus/AMOS (Palmer

et al., 2010), MetaVelvet-SL* (Cheng, 2015)

GENOME CHARACTERIZATION

1. Bacterial strain identification BLASTN (McGinnis and Madden, 2004), JSpeciesWS (Richter

et al., 2016), ANItools (Han et al., 2016), OrthoANI (Lee et al.,

2016), KmerFinder (Hasman et al., 2014), StrainSeeker (Roosaare

et al., 2017), MESH (Ondov et al., 2016), Kraken* (Wood and

Salzberg, 2014), MetaPhlAn* (Segata et al., 2012), QIIME2**

(Caporaso et al., 2010), MOTHUR** (Schloss et al., 2009),

MG-RAST** (Meyer et al., 2008)

PaPrBaG (Deneke et al., 2017), NBC (Rosen et al.,

2008), TACOA (Diaz et al., 2009), PhyloPythiaS+*

(McHardy et al., 2007; Gregor et al., 2016), BLCA**

(Gao et al., 2017), 16S Classifier** (Chaudhary et al.,

2015)

2. Bacterial genome annotation PROKKA (Seemann, 2014), RAST/myRAST (Overbeek et al.,

2014), MetaGeneAnnotator* (Noguchi et al., 2008), MetaGene*

(Noguchi et al., 2006), Tax4Fun** (Aßhauer et al., 2015)

Woods (Sharma et al., 2015), Orphelia* (Hoff et al.,

2009), MGC* (El Allali and Rose, 2013), MetaGUN*

(Liu et al., 2013), Meta-MFDL* (Chen et al., 2016)

3. Virulence gene detection VirulenceFinder (Joensen et al., 2014), PathogenFinder (Cosentino

et al., 2013)

BacFier (Iraola et al., 2012), PaPrBaG (Deneke et al.,

2017)

4. Antimicrobial resistance gene detection ResFinder (Zankari et al., 2012), RGI/CARD (Jia et al., 2017),

AMRFinder (Feldgarden et al., 2019)

DeepARG (Arango-Argoty et al., 2018), PATRIC

(Antonopoulos et al., 2017)

COMPARATIVE GENOMICS

1. Reference-based SNP methods CSI Phylogeny (Kaas et al., 2014), Lyve-SET (Katz et al., 2017),

CFSAN SNP Pipeline (Davis et al., 2015), SPANDx (Sarovich and

Price, 2014), SNVPhyl (Petkau et al., 2017)

2. Non-reference-based SNP analysis KSNP (Gardner et al., 2015)

3. Pangenome-based analysis Roary (Page et al., 2015), PanWeb (Pantoja et al., 2017), Pan-Seq

(Laing et al., 2010)

4. Core genome/whole-genome

multi-locus sequence typing (MLST)

EnteroBase (Alikhan et al., 2018), BIGSdb (Jolley and Maiden,

2010), chewBBACA (Silva et al., 2018)

BAPS/hierBAPS (Cheng et al., 2011, 2013)

PHYLOGENOMICS

RAxML (Stamatakis et al., 2005), FastTree (Price et al., 2009), CSI

Phylogeny (Kanamori et al., 2015), Lyve-SET (Katz et al., 2017),

PHYLIP (Shimada and Nishida, 2017), BEAST (Drummond and

Rambaut, 2007)

*The tool is dedicated to shotgun metagenomics; ** the tool dedicated to 16S metataxonomics.

metagenOmic data), which combines homology search thatconsiders also the secondary structure and properties of 16Sribosomal RNA genes to perform their de novo reconstruction(Yuan et al., 2015).

ML has been used in PERGA (Paired-End Reads GuidedAssembler) (Zhu et al., 2014) to determine the correctcontig extension. For this, the alogirthm constructs a decisionmodel, considers the avaialble information from paired-endreads such as different read overlap size and various branchfeatures, i.e., path weight, read coverage levels and gapsize. In addition, PERGA also detects tandem repeats withthe aim to resolve branches in the assembly graph andconstruct longer and more accurate contigs and scaffolds(Zhu et al., 2014). Minimus/AMOS (Palmer et al., 2010)contains a module that uses ML (C4.5 decision tree, NBand RF) in combination with features identified from priorsequencing projects and completed genomes to classify overlapsas true or false, by this improving the quality of thegenome assembly.

For shotgun metagenomics, ML-based strategies has beenproposed in order to pre-allocate (i.e., cluster) reads intosimilar groups before the assembly step, thus reducing theoverall computational complexity of the process (Cheng, 2015).Moreover, when assembling metagenomics data, the de Bruijngraph is usually decomposed into individual sub-graphs to buildan isolated genome; however, there are still the so called chimericnodes, i.e., those present in more than one sub-graph, whichneed to be identified and split apart (Afiahayati et al., 2015).For this, ML (SVM) has been applied, e.g., as implemented inMetaVelvet-SL (Afiahayati et al., 2015).

3. MACHINE LEARNING FOR MICROBIALGENOME CHARACTERIZATION

After assembly, the bacterial identity of the isolate usuallyneeds to be identified, followed by genome annotation andidentification of those genes that might be of clinical importance,




such as antimicrobial resistance and virulence genes. Forthis, genome characterization tools are being developed whichcompare the assembled contigs to several reference databases ofknown genes and reference genomes (Quainoo et al., 2017).

3.1. Bacterial Strain IdentificationIn this category, computational tools, which can assess bacterialidentity either directly from reads or from pre-assebled contigsare used (Quainoo et al., 2017). Current tools are often based ongenome-wide sequence similarity statistics (Ciufo et al., 2018).NCBI BLAST (the Basic Local Alignment Search Tool) is one ofthe most popular alignment tools and its variant BLASTN canbe used to identify species from contigs using the NucleotideCollection (nr/nt) database, which contains all the microbialsequences from the NCBI database (McGinnis and Madden,2004). However, for large-scale read mapping, BLAST may betoo slow (Deneke et al., 2017). Generally, this approach may failto detect novel species in cases when closely related genomesare not found in the reference databases (Deneke et al., 2017),which are known to be biased toward cultivable pathogenicbacteria (Farrell et al., 2018). Average Nucleotide Identity (ANI)(Clingenpeel et al., 2015) has been recently proposed as analternative metrics for the identification and classification ofbacterial species, calculated by performing several pair-wisecomparisons of all sequences shared between two given strains.This method is implemented within tools such as JSpeciesWS(Richter et al., 2016), ANItools (Han et al., 2016), and OrthoANI(Lee et al., 2016). Alternatively, composition-based methods suchas KmerFinder (Hasman et al., 2014) exist, which employ aprecomputed database compiled using 1,647 complete bacterialgenomes from the NCBI database divided into 16-mers. Givenan input file of unknown bacterial species, the program providesan overview of all k-mers that match all the templates in thedatabase (i.e., the “standard” method) or counts all the k-mersthat might originate from a particular strain (i.e., the “winnertakes it all” method; Hasman et al., 2014). StrainSeeker (Roosaareet al., 2017) starts with a Newick-format tree and derives a listof k-mers for each node in that tree. Thereafter, the observedvs. expected fractions of node-specific k-mers are being analyzedto determine each node’s presence in the input data (Roosaareet al., 2017). MESH (Ondov et al., 2016) is another k-merbased strain identification algorithm that extends the MinHashdimensionality-reduction technique by reducing large (sets of)sequences into small, representative sketches, which are thenused to infer global mutation distances.

For shotgun metagenomics, Kraken (Wood and Salzberg,2014) is a k-mer based approach, which tries to match 31-mersfrom the input data to a pre-computed database, by consideringall reference genomes in which they occur and then mappingthese 31-mers to the lowest common ancestor. MetaPhlAn(Segata et al., 2012) first collects all clade-specific marker genes,i.e., from strain to phylum, into a database, which it then utilizedfor the taxonomic classification of metagenomic shotgun data.

For 16S metataxonomics data, sequence alignment-basedapproaches are usually used to assign taxa (Chaudhary et al.,2015). For this, QIIME2 (Caporaso et al., 2010), MOTHUR(Schloss et al., 2009), and MG-RAST (Meyer et al., 2008)

are the most commonly used pipelines. Overall, the majorlimitations of the above approaches are the computationaltime requirements and dependence on the reference databases(Chaudhary et al., 2015).

To overcome these limitations, ML-based approaches havebeen proposed. NBC (Rosen et al., 2008) calculates k-merfrequency profiles of all publicly available microbial referencegenomes and uses these profiles to train a naive Bayesian classifierto identify the respective genome by any query fragment.TACOA (Diaz et al., 2009) achieves taxonomic classificationby combining the k-nearest neighbor algorithms with kernel-based ML strategies. Yet another ML-based approach, PaPrBaG(Pathogenicity Prediction for Bacterial Genomes), has beenrecently proposed, which, in addition to taxonomic classification,also aims to predict the pathogenic potential of the respectivestrains (Deneke et al., 2017).

For shotgun metagenomics, PhyloPythiaS+ (McHardy et al.,2007; Gregor et al., 2016) is a sequence composition-basedmethod that uses hierarchical structured-output by employing amulticlass support vector machine (SVM) classifier.

For 16S metataxonomics data, prediction-based MLapproaches for taxonomic classification have started to emerge,as opposed to homology-basedmethods (Chaudhary et al., 2015).For example, BLCA is a tool for taxonomic classification of 16SrRNA gene sequences, which combines sequence similarity tothe reference database with Bayesian posterior probabilities toweight the degree of sequence similarity of the query sequence toevery hit from the database (Gao et al., 2017). 16S Classifier is asimilar tool that deploys RF and is compatible with the QIIME2pipeline (Chaudhary et al., 2015).

3.2. Bacterial Genome AnnotationBacterial genome annotation tools explore which genes arecontained in the respective bacterial genome by retrieving therelevant features (i.e., coding regions and their putative products,non-coding RNAs and signal peptides) from raw reads orpre-assembled contigs (Seemann, 2014; Quainoo et al., 2017).PROKKA (Seemann, 2014) is a software suite unifying severalfeature prediction tools, such as Prodigal (Hyatt et al., 2010) forthe identification of coding sequences, RNAmmer (Lagesen et al.,2007), Aragorn (Laslett and Canback, 2004), and Infernal (Kolbeand Eddy, 2011) for the prediction of ribosomal, transfer andnon-coding RNA genes, respectively, as well as SignalP (Petersenet al., 2011) to identify signal leader peptides. RAST/myRAST(Overbeek et al., 2014) is another popular genome annotationtool, which uses a SEED k-mer-based annotation algorithm topredict coding sequences, as well as tRNAs and rRNAs.

For shotgun metagenomics, there are several model-basedapproaches, including MetaGeneAnnotator (Noguchi et al.,2008) or MetaGene (Noguchi et al., 2006), both using Markovchain models to identify genes.

However, the main limitation of these models is that theyrequire optimization of thousands of parameters, which limitstheir practical use (Zhang et al., 2017). Sequence similarity-based methods, on the other hand, are considered rather time-consuming and computationally demanding, especially whenapplied to shotgun metagenomic data. This poses a bottleneck




for efficient sequencing data analysis (Sharma et al., 2015).Moreover, RAST is known to have difficulties dealing with mixedor contaminated cultures, as its algorithm relies on closely relatedisolates (Quainoo et al., 2017). In addition, these methods areused to find genes with previously known homologous proteinsand cannot predict novel genes (Zhang et al., 2017).

Unfortunately, 16S metataxonomic data does not provide anyinformation on functional genes and proteins for the microbialcommunities being analyzed (Aßhauer et al., 2015); however,these can be predicted using pangenome-based approaches suchas Tax4Fun (Aßhauer et al., 2015).

Alternatively, ML (RF) and similarity-based (RAPsearch2)approaches have been combined in a tool called “Woods”(Sharma et al., 2015); however, it is currently restricted to theprediction of protein coding sequences only.

For shotgun metagenomics, several ML-based methods havebeen proposed, such as Orphelia (Hoff et al., 2009), MGC(El Allali and Rose, 2013), MetaGUN (Liu et al., 2013), andMeta-MFDL (Zhang et al., 2017), e.g., the latter using a deep stackingnetworks learning model and multiple genomic features (i.e.,the usage of monocodons and monoamino acids) for identifyinggenes from metagenomic fragments (Zhang et al., 2017).

3.3. Virulence Gene DetectionIn this part of the analysis, the aim is to explore whether thepreviously annotated genes infer virulence, i.e., some degreeof pathogenicity to the host (Quainoo et al., 2017). However,virulence gene detection does not necessarily have to follow thegenome annotation step. It can also be performed either usingreference database entries as BLAST queries against assembledgenomes or mapping raw reads against reference database entries(or any other collection of genes of interest). Also, predicted (butnot annotated) coding DNA (or predicted protein) sequences canbe screened for virulence gene content. Themost commonly usedreference database for virulence genes is the Virulence FactorDatabase (VFDB) (Chen et al., 2016), containing information on951 bacterial strains and 1,075 virulence factors (as of March2019), including different characteristics, such as whether avirulence factor is used in offensive or defensive actions. Recently,VFDB has been supplemented with VFanalyzer, a Web-basedtool that builds orthologous groups of genes using a querygenome and pre-analyzed reference genomes and then performssequence similarity searches among the VFDB gene collectionfor atypical and strain-specific virulence genes (https://doi.org/10.1093/nar/gky1080). Frequently used tools to predict virulencegenes from sequencing data include VirulenceFinder (Joensenet al., 2014), a Web-based tool that uses BLASTN (Camachoet al., 2009) and contains virulence markers for four microbes:Listeria, S. aureus, E. coli, and Enterococcus. Another Web-basedtool is PathogenFinder (Cosentino et al., 2013), which assumesthat bacterial pathogenicity (or lack of it) depends on groups ofproteins that are consistently found together in either pathogensor non-pathogens. PathogenFinder aims to identify such groupsof proteins.

Several ML-based approaches have been proposed forvirulence gene detection. VirulentPred (Garg and Gupta, 2008)is a bi-layer cascade SVM-based prediction method, where

the first layer classifiers are being trained using differentprotein sequence features, such as amino acid and dipeptidecomposition. The results from the first layer are then passed tothe second layer classifier, which utilizes sequence similarity anda BLAST database containing both virulence and non-virulencegenes. BacFier (Iraola et al., 2012) uses known pathogenicvs. non-pathogenic strains and their genetic features (e.g., thepresence or absence of different virulence-related genes) to trainML algorithms in predicting pathogenicity of input bacterialgenomes. Finally, as described above, PaPrBaG (Deneke et al.,2017) also aims to predict the pathogenic potential of microbialstrains by means of training on a large number of establishedpathogenic species in comparison with non-pathogenic bacteriaand their sequence features. PaPrBaG is a RF-based methodfor the assessment of the pathogenic potential of a set of readsbelonging to a single genome. It helps in the prediction of novel,unknown bacterial pathogens. PaPrBaG provides prediction incontrast with other approaches that discard many sequencingreads based on the low similarity to known reference genomes.

3.4. Antimicrobial Resistance GeneDetectionIn this step, computational analysis is used to explore whetherthe previously annotated bacterial genes infer antimicrobialresistance, i.e., the ability of microorganisms to grow despiteexposure to antimicrobial substances (Quainoo et al., 2017).However, again, the same is true as for virulence geneprediction—this step does not necessarily have to follow thegenome annotation step, e.g., it can be also conducted rightafter assembly. Frequently used tools for this purpose include aWeb-based tool ResFinder (Zankari et al., 2012) and RGI/CARD(Jia et al., 2017). Both perform homology-based resistomeprediction: ResFinder (Zankari et al., 2012) uses BLAST, whereasRGI/CARD (Jia et al., 2017) makes use of a manually curatedresource containing antimicrobial resistance genes, proteins andmutated sequences—CARD (Jia et al., 2017). Resently, NCBI hasdeveloped AMRFinder (Feldgarden et al., 2019) which utilizesthe NCBI’s curated AMR gene database - Bacterial AntimicrobialResistance Reference Gene Database-, currently including 4,579antimicrobial resistance gene proteins and over 560 hiddenMarkov models (HMMs).

ML approaches for the same task include DeepARG (Arango-Argoty et al., 2018), a deep learning approach using neuralnetworks and previously curated databases, such as CARD(Jia et al., 2017), for predicting antibiotic resistance genes andannotating them to 30 known antibiotic resistance categories,creating a manually curated database, DeepARG-DB. PATRIC(Antonopoulos et al., 2017) uses the genomes in its in-housedatabase and their antimicrobial resistance-related metadata,such as susceptibility or resistance to a given antibiotic, to buildAdaBoost (adaptive boosting) ML-based classifiers and predictthose regions within a bacterial genome that are associated withantimicrobial resistance (Davis et al., 2016). When a genomeis submitted to the PATRIC annotation service, these classifiersare used to predict if the organism is susceptible or resistant toan antibiotic. However, PATRIC is limited to identifying only


https://doi.org/10.1093/nar/gky1080https://doi.org/10.1093/nar/gky1080https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


genes encoding resistance to certain antibiotics (beta lactam,carbapenem, and methicillin) and in certain bacterial species.In this context, ML has also been applied to identify genomicfeatures possibly related to minimum inhibitory concentration(MIC) of an antibiotic, i.e., its lowest concentration preventingvisible growth of bacterium in vitro, e.g., for NontyphoidalSalmonella (Nguyen et al., 2019).

4. MACHINE LEARNING FOR MICROBIALCOMPARATIVE GENOMICS

After characterization of an individual genome is accomplished,the next step is to perform comparative genomics and detectrelatedness between strains, identify potentially clonal strainsand pinpoint the putative source of the outbreak (Brown et al.,2019). Bacterial species should be determined before performingcomparative genomic analyses, since most algorithms willperform better when closely related bacterial strains can beused. Comparative genomics methods can be largely dividedinto three groups: (i) reference/non-reference-based SNP-basedmethods, (ii) pangenome-based and (iii) core genome/whole-genome multilocus sequence typing (MLST).

4.1. Reference-Based SNP MethodsStandard strategies to identify genetic variation, which occursin a strain, usually focus on single nucleotide polymorphisms(SNPs). Raw reads are mapped to a perform better whenclosely related, high-quality reference genome, identifying SNPsas variations in relation to that reference genome. CSI Phylogeny(Kaas et al., 2014), Lyve-SET (Katz et al., 2017), CFSAN SNPPipeline (Davis et al., 2015), SPANDx (Sarovich and Price, 2014),and SNVPhyl (Petkau et al., 2017) include such pipelines. Inaddition, there are also tools such as Harvest/Parsnp (Treangenet al., 2014) that, instead of trying to performing whole-genomealignment, focus on constructing a core-genome alignment,i.e., identifying a set of orthologous sequence conserved inall aligned genomes. However, reference-based SNP methodsare generally recommended only if a high-quality referencegenome exists (Brown et al., 2019), when higher resolution isrequired than can be achieved using cgMLST/wgMLST, or whena cgMLST/wgMLST scheme is not available (Katz et al., 2017).

4.2. Reference-Free SNP AnalysisReference-Free SNP Analysis does not require alignment toa reference genome to identify SNPs. Such examples includekSNP (Gardner et al., 2015), a k-mer-based approach wherethe user provides the length of the flanking sequence includingthe SNP, i.e., the SNP is at the central base of the k-mer, andthe flanking (k-1)/2 bases on both sides of the SNP define thelocus. First, kSNP counts all k-mer oligos for each input genome.This is followed by several filtering steps: (i) the k-mer list isthen condensed so that counts reflect both occurrences on theforward and reverse strands; (ii) for raw reads, kSNP discardsk-mers that occur only once, as such singletons are likely tobe sequencing errors; (iii) for each genome, kSNP discards k-mers that have more than one central base variant for a givenlocus. Finally, kSNP merges and sorts all k-mers across all userprovided genomes and looks for SNP loci in themerged list. Then

it compares the SNP loci for each genome with the merged list toidentify the SNPs in each genome, reporting the locus and thecentral base, i.e., the SNP, for every genome containing that locus(Gardner et al., 2015).

4.3. Pangenome-Based AnalysisPangenome-based analysis classifies genes as the so called coregenes, found in all bacterial strains under comparison, and intoaccessory genes that can be found only in several but not allstrains (Page et al., 2015). Isolates are then clustered based ontheir accessory genome (Page et al., 2015). A well-known toolfor pangenome-based analysis is Roary (Page et al., 2015). First,it identifies orthologous genes by sequence comparison. This isfollowed by grouping of these genes into clusters. Finally, therelationships of the clusters are then represented using a graph,constructed based on the order in which their occur in theinput data (Page et al., 2015; Brown et al., 2019). Other tools forpangenome-based analysis include PanWeb (Pantoja et al., 2017)and Pan-Seq (Laing et al., 2010).

4.4. Core Genome/Whole-GenomeMulti-locus Sequence Typing (MLST)Core genome/whole-genome multi-locus sequence typing(MLST) are widely used methods for outbreak investigations,enabling standardized outbreak management protocols (Nadonet al., 2017; Brown et al., 2019). Conventional MLST usuallyuses only seven genes/loci to derive sequence types (STs), andis not always able to distinguish between outbreaks resultingfrom closely related bacterial variants (Pearce et al., 2018). Coregenome MLST (cgMLST) schemes extend the conventionalMLST, including genes/loci present in 95% to 99% of isolates,hence offering increased resolution to detect isolate-specificgenotypes, as well as novel transmission events (Nadon et al.,2017; Brown et al., 2019). If two strains display identical cgMLSTprofiles, these are being grouped into one cluster type (CT),which can be shared using dedicated databases (Quainoo et al.,2017). CgMLST is implemented within the Ridom SeqSphere+commercial software suite (JÃijnemann et al., 2013). However,it is also being utilized by EnteroBase (Alikhan et al., 2018),Bacterial Isolate Genome Sequence Database (BIGSdb) (Jolleyand Maiden, 2010) and chewBBACA (Silva et al., 2018). On theother hand, whole-genome MLST (wgMLST) further extendscgMLST, as it also considers the accessory genes to detectlineage-specific loci. This method is part of the BioNumerics(Applied Maths) software suite since version 7.5 (http://www.applied-maths.com/) and is also implemented within EnteroBase(Alikhan et al., 2018). For outbreak investigations, cgMLST ismore suited, as it uses species-specific nomenclature; however,wgMLST might offer higher resolution to discriminate outbreakstrains that form closely related clusters (Nadon et al., 2017;Brown et al., 2019). Of note, however, both methods stronglydepend on the availability of high-resolution isolate typingschemes (Pearce et al., 2018), which may not be available forlesser-studied foodborne pathogens, due to the lack of publiclyavailable WGS data (Carroll et al., 2019).

To the best of our knowledge, ML-based tools do not seemto have gained a lot of attention in comparative genomics.The Bayesian Analysis of Population Structure (BAPS)/hierBAPS


http://www.applied-maths.com/http://www.applied-maths.com/https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


(Cheng et al., 2011, 2013) tool seems to be the only ML-basedtool for comparative genomics. BAPS/hierBAPS was created byfirst collecting large data sets of multi-locus DNA sequence types(STs), as well as the respective metadata (e.g., host organism,serotype) from several MLST databases PubMLST (http://www.pubmlst.org). This data was then utilized to divide the availablepathogens into subsets of different evolutionary lineages orgeographically related sub-populations, as determined based onmolecular [dis]similarities within the database. Then a user-submitted set of bacterial isolates can be classified to one ofthese groups, using a Bayesian model-based ML algorithm.In addition, recently, several other studies have combinedcomparative genomics with ML approaches for the classificationof outbreak strains (Diaz et al., 2017) or source tracking duringoutbreaks (Buultjens et al., 2017; Zhang et al., 2019). Diaz et al.(2017) identified six distinct subtypes of genomes, as well astheir respective SNPs/loci, and trained RF to separate inputgenomes into the respective subtypes. Buultjens et al. (2017)used core genome variation and classification based on principalcomponents to identify genomic signatures specific to sourceof interest, which were further used to predict the origin ofinput isolates (Buultjens et al., 2017). Zhang et al. (2019) useda set of genetic features extracted from Salmonella Typhimuriumgenomes, inlcuding core genome SNPs, insertion/deletions andaccessory genes to train a RF classifier in discriminating isolatesfrom swine, bovine, poultry or wild bird sources. Wheeler et al.(2018) investigated genomic signatures related to host adaptationin Salmonella enterica. First, hidden Markov models were usedto identify patterns of sequence variation and their potentialfunctional consequences. Thereafter, RF was utilized to identifygenes that displayed differences between lineages with differentphenotypes (Wheeler et al., 2018). Sharma et al. (2014) usedMLST to differentiate isolates and categorize an unknown isolateas either representing a true infection or a likely contaminant.In particular, the seven genotypes derived from MLST were usedto train three different ML algorithms (SVM; Classification AndRegression Tree Analysis - CART; and a Naive Nearest-NeighborClassifier) to segregate isolates of known class (i.e., pathogen orlikely contaminant) on the basis of their alleles, which were thenused to classify an unknown isolate by its MLST allele profile.

5. MACHINE LEARNING FOR THEINFERENCE OF MICROBIALPHYLOGENOMICS

Finally, comparison tools can be used for the inference ofmicrobial phylogenomics of pathogenic isolates and generatedetailed networks reflecting the transmission events of outbreakstrains between different patients (Quainoo et al., 2017). Inparticular, phylogenomics can reveal whether two isolates arenearly identical or only distantly related and which mightrepresent the initial outbreak source strain (Quainoo et al., 2017).Maximum likelihood is frequently applied when characterizingpathogens from foodborne outbreaks. RAxML (RandomizedAxelerated Maximum Likelihood) (Stamatakis et al., 2005) andFastTree (Price et al., 2009) are two maximum likelihood based

phylogenomics estimators, which work by first constructing aninitial tree, which is then further refined in several optimizationsteps and tree rearrangements to increase the likelihoodthat the respective tree reflects the evolutionary relationshipsof the input sequences. These software packages are oftenincluded in the genome comparison pipelines mentioned inthe previous chapter such as CSI Phylogeny (Kaas et al.,2014) and Lyve-SET (Katz et al., 2017) for streamlinedproduction of actionable results. Alternatively, distance matrix-based methods such as neighbor joining (Saitou and Nei,1987) (e.g., part of the PHYLIP Shimada and Nishida, 2017package) as well as Bayesian analysis-basedmethods (e.g., BEASTDrummond and Rambaut, 2007) have been proposed to studymicrobial phylogenomics.

Most recently, Suvorov et al. (2019) has proposed an approachthat uses convolutional neural networks (CNNs) for phylogeneticinference. In particular, CNNs are being trained to extractphylogenetic signal from a multiple sequence alignment, whichis then used to reconstruct and discriminate alternative treetopologies. Of note, however, this study used an alignment of onlyfour sequences.

6. CONCLUSIONS

Over the last years, several ML-based tools have been developedfor different steps of bacterial WGS analysis. However, someareas of bacterial bioinformatics (i.e., genome assembly andstrain identification) have seen more development than others(i.e., phylogeny estimation). Overall, AI and its sub-disciplineML could lead to actionable knowledge in diverse rangesof sectors, where multiple complex challenges need to beaddressed, including the outbreak investigations of foodbornepathogens and antimicrobial resistance (Gibbs, 2014; Quainooet al., 2017; Ching et al., 2018), considering that WGS mayreplace conventional analysis methods already in the nearfuture (Quainoo et al., 2017). In this scenario, the success ofoutbreak investigations will largely depend on how fast andaccurate WGS data can be produced and analyzed (Quainooet al., 2017). ML-based algorithms could further speed-upsuch investigations, especially as the number of completemicrobial genomes in NCBI RefSeq (http://www.ncbi.nlm.nih.gov/genome) is rapidly growing (Tatusova et al., 2015), providinga valuable resource for training ML classifiers. However, evenif substantially improving the accuracy and speed of WGSalgorithms, a number of limitations still need to be overcomein order to fully utilize the power of ML for outbreakscreenings. WGS analysis tools often rely on sequence similarityand hence strongly depend on reference databases (Denekeet al., 2017; Zhang et al., 2017). Moreover, such methods arerather time-consuming and computationally demanding, thusrepresenting a bottleneck for efficient sequence data analysis(Sharma et al., 2015). ML algorithms could potentially increasethe accuracy and speed of clinically and epidemiologicallyrelevant predictions (Farrell et al., 2018). However, to yieldaccurate predictions, besides the choice of the most appropriatealgorithm and a set of well-defined inputs and outputs of


http://www.pubmlst.orghttp://www.pubmlst.orghttp://www.ncbi.nlm.nih.gov/genomehttp://www.ncbi.nlm.nih.gov/genomehttps://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


interest, ML-based strategies generally require large amountsof high-quality training data (Baker et al., 2018). This presentsa limitation, as currently microbial genome databases areknown to be biased toward cultivable pathogenic bacteria. Thecurrent lack of large and comprehensive databases can beconsidered as the key bottleneck for the application of MLmethods (Farrell et al., 2018). Hence, future improvementscan be expected to come from better data curation andcollection, in addition to development of new and improvedclassification algorithms (Farrell et al., 2018). Therefore, WGSdata collection must be done in parallel with comprehensiveand standartized metadata collection such as phenotypicprofiling using traditional microbiology methods for isolatecharacterization (e.g., phenotypic profiling of antimicrobialresistance) (Maurer et al., 2017).

Currently, sequencing of bacterial genomes is mostlyperformed on Illumina instruments, producing relativelyshort reads with limited resolution of low-complexity regions(Quainoo et al., 2017). Alternatively, ultra-long read technologiessuch as ONT (https://nanoporetech.com/) and PacBio SMRT(https://www.pacb.com/smrt-science/smrt-sequencing/) areincreasingly being used to obtain complete microbial genomes.However, both technologies are still three and almost seven timesmore expensive in comparison to Illumina short-read sequencing(Brown et al., 2017; Sekse et al., 2017; Nicola De Maio, 2019).Moreover, both technologies still display rather high error rates(Mahmoud et al., 2017), which makes themmore suitable for gapclosure in draft genomes using hybrid methods (Quainoo et al.,2017). Hence, error-profile-aware ML-algorithms implementinghybrid strategies that make use of more accurate short reads inconjunction with ultra-long reads may need to be considered forfuture applications.

The selection of a harmonized bioinformatics strategy orpipeline that would perform consistently across outbreakinvestigation situations around the world, reaching consensuson desired standards represents another challenge for theroutine implementation of WGS analysis (Quainoo et al.,2017). Especially, considering that the numbers of commercialanalysis software platforms, as well as open-source, application-specific analysis tools are increasing, a rigorous assessment andbenchmarking of their quality is urgently needed (Quainooet al., 2017). This would also be a prerequisite for a systematiccomparison between ML-based vs. conventional methods.Nevertheless, in order to perform such comparisons on aglobal scale, WGS data storage and sharing would be ofutmost importance. Although technically feasible, this willrequire us to solve several issues of ownership and dataprivacy, making sure that these are being adequately protected(Quainoo et al., 2017).

AUTHOR CONTRIBUTIONS

BV wrote the manuscript. IM, LG-I, and JK participated inrevising and editing the manuscript. All authors have read andapproved the final version of the manuscript.

FUNDING

This research was funded by the ERDF and state budget co-financed project No. 1.1.1.1/16/A/258 “Development and theapplication of innovative instrumental analytical methods forthe combined determination of a wide range of chemical andbiological contaminants in support of the bio-economy in thepriority sectors of economy”.

REFERENCES

Afiahayati, Sato, K., and Sakakibara, Y. (2015). Metavelvet-sl: an extension ofthe velvet assembler to a de novo metagenomic assembler utilizing supervisedlearning. DNA Res. 22, 69–77. doi: 10.1093/dnares/dsu041

Afify, H. M., and Al-Masni, M. A. (2018). Taxonomy metagenomic analysis formicrobial sequences in three domains system via machine learning approaches.Inform. Med. Unlocked 13, 151–157. doi: 10.1016/j.imu.2018.05.004

Alikhan, N.-F., Zhou, Z., Sergeant, M. J., and Achtman, M. (2018). A genomicoverview of the population structure of salmonella. PLoS Genet. 14:e1007261.doi: 10.1371/journal.pgen.1007261

Alkema, W., Boekhorst, J., Wels, M., and van Hijum, S. A. F. T. (2016). Microbialbioinformatics for food safety and production. Brief. Bioinformat. 17, 283–292.doi: 10.1093/bib/bbv034

Andersen, S. C., and Hoorfar, J. (2018). Surveillance of foodborne pathogens:Towards diagnostic metagenomics of fecal samples. Genes 9:E14.doi: 10.3390/genes9010014

Antonopoulos, D. A., Assaf, R., Aziz, R. K., Brettin, T., Bun, C., Conrad, N., et al.(2017). Patric as a unique resource for studying antimicrobial resistance. Brief.Bioinform. doi: 10.1093/bib/bbx083. [Epub ahead of print].

Arango-Argoty, G., Garner, E., Pruden, A., Heath, L. S., Vikesland, P.,and Zhang, L. (2018). Deeparg: a deep learning approach for predictingantibiotic resistance genes from metagenomic data. Microbiome 6:23.doi: 10.1186/s40168-018-0401-z

Aßhauer, K. P., Wemheuer, B., Daniel, R., and Meinicke, P. (2015). Tax4fun:predicting functional profiles from metagenomic 16s rrna data. Bioinformatics31, 2882–2884. doi: 10.1093/bioinformatics/btv287

Baker, R. E., Peña, J.-M., Jayamohan, J., and Jérusalem, A. (2018).Mechanistic models versus machine learning, a fight worth fighting forthe biological community? Biol. Lett. 14:20170660. doi: 10.1098/rsbl.2017.0660

Bankevich, A., Nurk, S., Antipov, D., Gurevich, A. A., Dvorkin, M., Kulikov,A. S., et al. (2012). Spades: a new genome assembly algorithm andits applications to single-cell sequencing. J. Comput. Biol. 19, 455–477.doi: 10.1089/cmb.2012.0021

Bergholz, T. M., Moreno Switt, A. I., and Wiedmann, M. (2014). Omicsapproaches in food safety: fulfilling the promise? TrendsMicrobiol. 22, 275–281.doi: 10.1016/j.tim.2014.01.006

Boisvert, S., Laviolette, F., and Corbeil, J. (2010). Ray: simultaneous assembly ofreads from a mix of high-throughput sequencing technologies. J. Comput. Biol.17, 1519–1533. doi: 10.1089/cmb.2009.0238

Boser, B. E., Guyon, I. M., and Vapnik, V. N. (1992). “A training algorithm foroptimal margin classifiers,” in Proceedings of the Fifth Annual Workshop onComputational Learning Theory (Pittsburgh, PA: ACM), 144–152.

Breiman, L. (1996). Bagging predictors.Mach. Learn. 24, 123–140.Brown, B. L., Watson, M., Minot, S. S., Rivera, M. C., and Franklin, R. B. (2017).

MinIONTM nanopore sequencing of environmental metagenomes: a syntheticapproach. GigaScience 6, 1–10. doi: 10.1093/gigascience/gix007

Brown, E., Dessai, U., McGarry, S., and Gerner-Smidt, P. (2019). Useof whole-genome sequencing for food safety and public health in theunited states. Foodborne Pathogens Dis. 16, 441–450. doi: 10.1089/fpd.2019.2662

Buultjens, A. H., Chua, K. Y. L., Baines, S. L., Kwong, J., Gao, W., Cutcher, Z.,et al. (2017). A supervised statistical learning approach for accurate Legionella


https://nanoporetech.com/https://www.pacb.com/smrt-science/smrt-sequencing/https://doi.org/10.1093/dnares/dsu041https://doi.org/10.1016/j.imu.2018.05.004https://doi.org/10.1371/journal.pgen.1007261https://doi.org/10.1093/bib/bbv034https://doi.org/10.3390/genes9010014https://doi.org/10.1093/bib/bbx083https://doi.org/10.1186/s40168-018-0401-zhttps://doi.org/10.1093/bioinformatics/btv287https://doi.org/10.1098/rsbl.2017.0660https://doi.org/10.1089/cmb.2012.0021https://doi.org/10.1016/j.tim.2014.01.006https://doi.org/10.1089/cmb.2009.0238https://doi.org/10.1093/gigascience/gix007https://doi.org/10.1089/fpd.2019.2662https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


pneumophila source attribution during outbreaks.Appl. Environ. Microbiol. 83:e01482-17. doi: 10.1128/AEM.01482-17

Camacho, C., Coulouris, G., Avagyan, V., Ma, N., Papadopoulos, J., Bealer, K.,et al. (2009). Blast+: architecture and applications. BMC Bioinform. 10:421.doi: 10.1186/1471-2105-10-421

Caporaso, J. G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman,F. D., Costello, E. K., et al. (2010). Qiime allows analysis of high-throughput community sequencing data. Nat. Methods 7, 335–336.doi: 10.1038/nmeth.f.303

Carroll, L. M., Wiedmann, M., Mukherjee, M., Nicholas, D. C., Mingle,L. A., Dumas, N. B., et al. (2019). Characterization of emetic anddiarrheal bacillus cereus strains from a 2016 foodborne outbreakusing whole-genome sequencing: addressing the microbiological,epidemiological, and bioinformatic challenges. Front. Microbiol. 10:144.doi: 10.3389/fmicb.2019.00144

CDC (2013). Antibiotic Resistance Threats in the United States, 2013. Technicalreport, CDC. Atlanta, GA.

Chaudhary, N., Sharma, A. K., Agarwal, P., Gupta, A., and Sharma, V. K. (2015).16s classifier: a tool for fast and accurate taxonomic classification of 16srrna hypervariable regions in metagenomic datasets. PLoS ONE 10:e0116106.doi: 10.1371/journal.pone.0116106

Chen, L., Zheng, D., Liu, B., Yang, J., and Jin, Q. (2016). Vfdb 2016: hierarchicaland refined dataset for big data analysis–10 years on. Nucleic Acids Res. 44,D694–D697. doi: 10.1093/nar/gkv1239

Cheng, L. (2015). A Machine Learning Approach to DNA Shotgun SequenceAssembly. PhD thesis, University of the Witwatersrand. Johannesburg,South Africa.

Cheng, L., Connor, T. R., Aanensen, D. M., Spratt, B. G., and Corander, J.(2011). Bayesian semi-supervised classification of bacterial samples usingMLST databases. BMC Bioinform. 12:302. doi: 10.1186/1471-2105-12-302

Cheng, L., Connor, T. R., Sirén, J., Aanensen, D. M., and Corander, J. (2013).Hierarchical and spatially explicit clustering of dna sequences with bapssoftware.Mol. Biol. Evol. 30, 1224–1228. doi: 10.1093/molbev/mst028

Ching, T., Himmelstein, D. S., Beaulieu-Jones, B. K., Kalinin, A. A., Do, B. T., Way,G. P., et al. (2018). Opportunities and obstacles for deep learning in biology andmedicine. J. R. Soc. Interface 15: 20170387. doi: 10.1098/rsif.2017.0387

Ciufo, S., Kannan, S., Sharma, S., Badretdin, A., Clark, K., Turner, S., et al.(2018). Using average nucleotide identity to improve taxonomic assignments inprokaryotic genomes at the NCBI. Int. J. Syst. Evol. Microbiol. 68, 2386–2392.doi: 10.1099/ijsem.0.002809

Clingenpeel, S., Clum, A., Schwientek, P., Rinke, C., and Woyke, T. (2015).Reconstructing each cell’s genome within complex microbial communities’dream or reality? Front. Microbiol. 5:771. doi: 10.3389/fmicb.2014.00771

Cosentino, S., Voldby Larsen, M., Møller Aarestrup, F., and Lund, O. (2013).Pathogenfinder–distinguishing friend from foe using bacterial whole genomesequence data. PLoS ONE 8:e77302. doi: 10.1371/journal.pone.0077302

Davis, J. J., Boisvert, S., Brettin, T., Kenyon, R. W., Mao, C., Olson, R., et al.(2016). Antimicrobial resistance prediction in patric and rast. Sci. Rep. 6:27930.doi: 10.1038/srep27930

Davis, S., Pettengill, J. B., Luo, Y., Payne, J., Shpuntoff, A., Rand, H., et al.(2015). Cfsan snp pipeline: an automated method for constructing snpmatrices from next-generation sequence data. PeerJ Comput. Sci. 1:e20.doi: 10.7717/peerj-cs.20

Deneke, C., Rentzsch, R., and Renard, B. Y. (2017). Paprbag: a machine learningapproach for the detection of novel pathogens fromNGS data. Sci. Rep. 7:39194.doi: 10.1038/srep39194

Devroye, L., Györfi, L., and Lugosi, G. (2013). A Probabilistic Theory of PatternRecognition, Vol 31. Springer Science & Business Media.

Diaz, M. H., Desai, H. P., Morrison, S. S., Benitez, A. J., Wolff, B. J., Caravas, J., etal. (2017). Comprehensive bioinformatics analysis of mycoplasma pneumoniaegenomes to investigate underlying population structure and type-specificdeterminants. PLoS ONE 12:e0174701. doi: 10.1371/journal.pone.0174701

Diaz, N. N., Krause, L., Goesmann, A., Niehaus, K., and Nattkemper, T. W.(2009). Tacoa: taxonomic classification of environmental genomic fragmentsusing a kernelized nearest neighbor approach. BMC Bioinform. 10:56.doi: 10.1186/1471-2105-10-56

Drummond, A. J., and Rambaut, A. (2007). Beast: bayesian evolutionary analysisby sampling trees. BMC Evol. Biol. 7:214. doi: 10.1186/1471-2148-7-214

Dutilh, B. E., Backus, L., Edwards, R. A., Wels, M., Bayjanov, J. R., and van Hijum,S. A. F. T. (2013). Explaining microbial phenotypes on a genomic scale: Gwasfor microbes. Brief. Funct. Genom. 12, 366–380. doi: 10.1093/bfgp/elt008

EFSA (2015). The european union summary report on trends and sources ofzoonoses, zoonotic agents and food-borne outbreaks in (2014). EFSA J. 13:191.doi: 10.2903/j.efsa.2015.4329

El Allali, A., and Rose, J. R. (2013). Mgc: a metagenomic gene caller. BMCBioinform. 14 (Suppl. 9):S6. doi: 10.1186/1471-2105-14-S9-S6

Farrell, F., Soyer, O. S., and Quince, C. (2018). Machine learning based predictionof functional capabilities in metagenomically assembled microbial genomes.bioRxiv. doi: 10.1101/307157

Feldgarden, M., Brover, V., Haft, D. H., Prasad, A. B., Slotta, D. J., Tolstoy, I., etal. (2019). Using the ncbi amrfinder tool to determine antimicrobial resistancegenotype-phenotype correlations within a collection of narms isolates. bioRxiv.doi: 10.1101/550707

Gao, X., Lin, H., Revanna, K., and Dong, Q. (2017). A bayesian taxonomicclassification method for 16s rRNA gene sequences with improved species-levelaccuracy. BMC Bioinform. 18:247. doi: 10.1186/s12859-017-1670-4

Gardner, S. N., Slezak, T., and Hall, B. G. (2015). ksnp3.0: Snp detection andphylogenetic analysis of genomes without genome alignment or referencegenome. Bioinformatics 31, 2877–2878. doi: 10.1093/bioinformatics/btv271

Garg, A., and Gupta, D. (2008). Virulentpred: a SVM based predictionmethod for virulent proteins in bacterial pathogens. BMC Bioinform. 9:62.doi: 10.1186/1471-2105-9-62

Gibbs, E. P. J. (2014). The evolution of one health: a decade of progress andchallenges for the future. Veter. Record 174, 85–91. doi: 10.1136/vr.g143

Gregor, I., Dröge, J., Schirmer, M., Quince, C., and McHardy, A. C. (2016).Phylopythias+: a self-training method for the rapid reconstructionof low-ranking taxonomic bins from metagenomes. PeerJ 4:e1603.doi: 10.7717/peerj.1603

Han, N., Qiang, Y., and Zhang, W. (2016). Anitools web: a web tool for fastgenome comparison within multiple bacterial strains. Database 2016: baw084.doi: 10.1093/database/baw084

Hartigan, J. A., and Wong, M. A. (1979). Algorithm AS 136: a k-means clusteringalgorithm. J. R. Statist. Soc. Ser. C 28, 100–108.

Hasman, H., Saputra, D., Sicheritz-Ponten, T., Lund, O., Svendsen, C. A., Frimodt-MÃÿller, N., et al. (2014). Rapid whole-genome sequencing for detection andcharacterization of microorganisms directly from clinical samples. J. Clin.Microbiol. 52, 139–146. doi: 10.1128/JCM.02452-13

Hendriksen, R. S., Pedersen, S. K., Leekitcharoenphon, P., Malorny, B., Borowiak,M., Battisti, A., et al. (2018). Final report of engage-establishing next generationsequencing ability for genomic analysis in europe. EFSA Supp. Public. 15:1431E.doi: 10.2903/sp.efsa.2018.EN-1431

Her, H.-L., and Wu, Y.-W. (2018). A pan-genome-based machine learningapproach for predicting antimicrobial resistance activities of the Escherichia colistrains. Bioinformatics 34, i89–i95. doi: 10.1093/bioinformatics/bty276

Hoff, K. J., Lingner, T., Meinicke, P., and Tech, M. (2009). Orphelia: predictinggenes in metagenomic sequencing reads. Nucleic Acids Res. 37, W101–W105.doi: 10.1093/nar/gkp327

Hotelling, H. (1933). Analysis of a complex of statistical variables into principalcomponents. J. Educ. Psychol. 24:417.

Hyatt, D., Chen, G.-L., Locascio, P. F., Land, M. L., Larimer, F. W., and Hauser,L. J. (2010). Prodigal: prokaryotic gene recognition and translation initiationsite identification. BMC Bioinform. 11:119. doi: 10.1186/1471-2105-11-119

Iraola, G., Vazquez, G., Spangenberg, L., and Naya, H. (2012). Reduced set ofvirulence genes allows high accuracy prediction of bacterial pathogenicity inhumans. PLoS ONE 7:e42144. doi: 10.1371/journal.pone.0042144

Jia, B., Raphenya, A. R., Alcock, B., Waglechner, N., Guo, P., Tsang, K. K.,et al. (2017). Card 2017: expansion and model-centric curation of thecomprehensive antibiotic resistance database. Nucleic Acids Res. 45, D566–D573. doi: 10.1093/nar/gkw1004

Joensen, K. G., Scheutz, F., Lund, O., Hasman, H., Kaas, R. S., Nielsen, E. M., et al.(2014). Real-time whole-genome sequencing for routine typing, surveillance,and outbreak detection of verotoxigenic Escherichia coli. J. Clin. Microbiol. 52,1501–1510. doi: 10.1128/JCM.03617-13

Jolley, K. A., and Maiden, M. C. (2010). Bigsdb: scalable analysis ofbacterial genome variation at the population level. BMC Bioinform. 11:595.doi: 10.1186/1471-2105-11-595


https://doi.org/10.1128/AEM.01482-17https://doi.org/10.1186/1471-2105-10-421https://doi.org/10.1038/nmeth.f.303https://doi.org/10.3389/fmicb.2019.00144https://doi.org/10.1371/journal.pone.0116106https://doi.org/10.1093/nar/gkv1239https://doi.org/10.1186/1471-2105-12-302https://doi.org/10.1093/molbev/mst028https://doi.org/10.1098/rsif.2017.0387https://doi.org/10.1099/ijsem.0.002809https://doi.org/10.3389/fmicb.2014.00771https://doi.org/10.1371/journal.pone.0077302https://doi.org/10.1038/srep27930https://doi.org/10.7717/peerj-cs.20https://doi.org/10.1038/srep39194https://doi.org/10.1371/journal.pone.0174701https://doi.org/10.1186/1471-2105-10-56https://doi.org/10.1186/1471-2148-7-214https://doi.org/10.1093/bfgp/elt008https://doi.org/10.2903/j.efsa.2015.4329https://doi.org/10.1186/1471-2105-14-S9-S6https://doi.org/10.1101/307157https://doi.org/10.1101/550707https://doi.org/10.1186/s12859-017-1670-4https://doi.org/10.1093/bioinformatics/btv271https://doi.org/10.1186/1471-2105-9-62https://doi.org/10.1136/vr.g143https://doi.org/10.7717/peerj.1603https://doi.org/10.1093/database/baw084https://doi.org/10.1128/JCM.02452-13https://doi.org/10.2903/sp.efsa.2018.EN-1431https://doi.org/10.1093/bioinformatics/bty276https://doi.org/10.1093/nar/gkp327https://doi.org/10.1186/1471-2105-11-119https://doi.org/10.1371/journal.pone.0042144https://doi.org/10.1093/nar/gkw1004https://doi.org/10.1128/JCM.03617-13https://doi.org/10.1186/1471-2105-11-595https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


Jünemann, S., Sedlazeck, F. J., Prior, K., Albersmeier, A., John, U., Kalinowski,J., et al. (2013). Updating benchtop sequencing performance comparison. Nat.Biotechn. 31, 294–296. doi: 10.1038/nbt.2522

Kaas, R. S., Leekitcharoenphon, P., Aarestrup, F. M., and Lund, O. (2014). Solvingthe problem of comparing whole bacterial genomes across different sequencingplatforms. PLoS ONE 9:e104984. doi: 10.1371/journal.pone.0104984

Kanamori, H., Parobek, C. M., Weber, D. J., van Duin, D., Rutala, W. A., Cairns,B. A., et al. (2015). Next-generation sequencing and comparative analysis ofsequential outbreaks caused by multidrug-resistant acinetobacter baumannii ata large academic burn center. Antimicrob. Agents. Chemother. 60, 1249–1257.doi: 10.1128/AAC.02014-15

Katz, L. S., Griswold, T., Williams-Newkirk, A. J., Wagner, D., Petkau, A., Sieffert,C., et al. (2017). A comparative analysis of the lyve-set phylogenomics pipelinefor genomic epidemiology of foodborne pathogens. Front. Microbiol. 8:375.doi: 10.3389/fmicb.2017.00375

Kolbe, D. L., and Eddy, S. R. (2011). Fast filtering for rna homology search.Bioinformatics 27, 3102–3109. doi: 10.1093/bioinformatics/btr545

Koren, S., Walenz, B. P., Berlin, K., Miller, J. R., Bergman, N. H., and Phillippy,A. M. (2017). Canu: scalable and accurate long-read assembly via adaptive,javax.xml.bind.jaxbelement@19c8c323, -mer weighting and repeat separation.Genome Res. 27, 722–736. doi: 10.1101/gr.215087.116

Kruse, R., Borgelt, C., Braune, C., Mostaghim, S., and Steinbrecher, M.(2016). Computational Intelligence: A Methodological Introduction. Heidelberg:Springer.

Lagesen, K., Hallin, P., Rødland, E. A., Staerfeldt, H.-H., Rognes, T., and Ussery,D. W. (2007). Rnammer: consistent and rapid annotation of ribosomal rnagenes. Nucleic Acids Res. 35, 3100–3108. doi: 10.1093/nar/gkm160

Lai, K., Twine, N., O’Brien, A., Guo, Y., and Bauer, D. (2016). “Artificialintelligence and machine learning in bioinformatics,” in Encyclopedia ofBioinformatics and Computational Biology, ed M. Gribskov (Elsevier).doi: 10.1016/B978-0-12-809633-8.20325-7

Laing, C., Buchanan, C., Taboada, E. N., Zhang, Y., Kropinski, A., Villegas, A., etal. (2010). Pan-genome sequence analysis using panseq: an online tool for therapid analysis of core and accessory genomic regions. BMC Bioinform. 11:461.doi: 10.1186/1471-2105-11-461

Laslett, D., and Canback, B. (2004). Aragorn, a program to detect trna genesand tmrna genes in nucleotide sequences. Nucleic Acids Res. 32, 11–16.doi: 10.1093/nar/gkh152

Lee, I., Ouk Kim, Y., Park, S.-C., and Chun, J. (2016). Orthoani: An improvedalgorithm and software for calculating average nucleotide identity. Int. J. Syst.Evol. Microbiol. 66, 1100–1103. doi: 10.1099/ijsem.0.000760

Li, H. (2016). Minimap and miniasm: fast mapping and de novoassembly for noisy long sequences. Bioinformatics 32, 2103–2110.doi: 10.1093/bioinformatics/btw152

Li, L.-G., Yin, X., and Zhang, T. (2018). Tracking antibiotic resistancegene pollution from different sources using machine-learning classification.Microbiome 6:93. doi: 10.1186/s40168-018-0480-x

Liu, Y., Guo, J., Hu, G., and Zhu, H. (2013). Gene prediction in metagenomicfragments based on the svm algorithm. BMC Bioinform. 14 (Suppl. 5):S12.doi: 10.1186/1471-2105-14-S5-S12

Llarena, A.-K., Ribeiro-Gonçalves, B. F., Nuno Silva, D., Halkilahti, J., Machado,M. P., Da Silva, M. S., et al. (2018). Innuendo: A cross-sectoral platform forthe integration of genomics in the surveillance of food-borne pathogens. EFSASupp. Public. 15:1498E. doi: 10.2903/sp.efsa.2018.EN-1498

Lupolova, N., Dallman, T. J., Holden, N. J., and Gally, D. L. (2017).Patchy promiscuity: machine learning applied to predict the host specificityof salmonella enterica and Escherichia coli. Microb. Genom. 3:e000135.doi: 10.1099/mgen.0.000135

Mahmoud, M., Zywicki, M., Twardowski, T., and Karlowski, W. M. (2017).Efficiency of pacbio long read correction by 2nd generation illuminasequencing. Genomics 111, 43–49. doi: 10.1016/j.ygeno.2017.12.011

Maurer, F. P., Christner, M., Hentschke, M., and Rohde, H. (2017). Advancesin rapid identification and susceptibility testing of bacteria in the clinicalmicrobiology laboratory: implications for patient care and antimicrobialstewardship programs. Infect. Dis. Rep. 9:6839. doi: 10.4081/idr.2017.6839

McGinnis, S., and Madden, T. L. (2004). Blast: at the core of a powerful anddiverse set of sequence analysis tools. Nucleic Acids Res. 32, W20–W25.doi: 10.1093/nar/gkh435

McHardy, A. C., Martín, H. G., Tsirigos, A., Hugenholtz, P., and Rigoutsos, I.(2007). Accurate phylogenetic classification of variable-length dna fragments.Nat. Methods 4, 63–72. doi: 10.1038/nmeth976

Meyer, F., Paarmann, D., D’Souza, M., Olson, R., Glass, E. M., Kubal, M., et al.(2008). The metagenomics rast server - a public resource for the automaticphylogenetic and functional analysis of metagenomes. BMC Bioinform. 9:386.doi: 10.1186/1471-2105-9-386

Moran-Gilad, J. (2017). Whole genome sequencing (wgs) for food-bornepathogen surveillance and control - taking the pulse. Euro Surveill. 22:30547.doi: 10.2807/156

Nadon, C., Van Walle, I., Gerner-Smidt, P., Campos, J., Chinen, I., Concepcion-Acevedo, J., et al. (2017). Pulsenet international: vision for the implementationof whole genome sequencing (wgs) for global food-borne disease surveillance.Euro Surveill. 22: 30544. doi: 10.2807/1560-7917.ES.2017.22.23.30544

Nguyen, M., Long, S. W., McDermott, P. F., Olsen, R. J., Olson, R., Stevens, R. L., etal. (2019). Using machine learning to predict antimicrobial mics and associatedgenomic features for nontyphoidal Salmonella. J. Clin.Microbiol. 57: e01260-18.doi: 10.1128/JCM.01260-18

Nicola De Maio, Liam P. Shaw, A. H. S. G. (2019). Comparison of long-readsequencing technologies in the hybrid assembly of complex bacterial genomes.bioRxiv doi: 10.1101/530824

Noguchi, H., Park, J., and Takagi, T. (2006). Metagene: prokaryotic gene findingfrom environmental genome shotgun sequences. Nucleic Acids Res. 34, 5623–5630. doi: 10.1093/nar/gkl723

Noguchi, H., Taniguchi, T., and Itoh, T. (2008). Metageneannotator: detectingspecies-specific patterns of ribosomal binding site for precise gene predictionin anonymous prokaryotic and phage genomes. DNA Res. 15, 387–396.doi: 10.1093/dnares/dsn027

Ondov, B. D., Treangen, T. J., Melsted, P., Mallonee, A. B., Bergman,N. H., Koren, S., and Phillippy, A. M. (2016). Mash: fast genome andmetagenome distance estimation using minhash. Genome Biol. 17:132.doi: 10.1186/s13059-016-0997-x

Overbeek, R., Olson, R., Pusch, G. D., Olsen, G. J., Davis, J. J., Disz, T., et al. (2014).The seed and the rapid annotation of microbial genomes using subsystemstechnology (rast). Nucleic Acids Res. 42, D206–D214. doi: 10.1093/nar/gkt1226

Page, A. J., Cummins, C. A., Hunt, M., Wong, V. K., Reuter, S., Holden, M.T. G., et al. (2015). Roary: rapid large-scale prokaryote pan genome analysis.Bioinformatics 31, 3691–3693. doi: 10.1093/bioinformatics/btv421

Palmer, L. E., Dejori, M., Bolanos, R., and Fasulo, D. (2010). Improving de novosequence assembly using machine learning and comparative genomics foroverlap correction. BMC Bioinform. 11:33. doi: 10.1186/1471-2105-11-33

Pantoja, Y., Pinheiro, K., Veras, A., AraÃžjo, F., Lopes de Sousa, A., GuimarÃčes,L. C., et al. (2017). Panweb: A web interface for pan-genomic analysis. PLoSONE 12:e0178154. doi: 10.1371/journal.pone.0178154

Pearce, M. E., Alikhan, N.-F., Dallman, T. J., Zhou, Z., Grant, K., and Maiden, M.C. J. (2018). Comparative analysis of core genome mlst and snp typing withina european salmonella serovar enteritidis outbreak. Int. J. Food Microbiol. 274,1–11. doi: 10.1016/j.ijfoodmicro.2018.02.023

Peng, Y., Leung, H. C. M., Yiu, S. M., and Chin, F. Y. L. (2012). Idba-ud: a de novoassembler for single-cell and metagenomic sequencing data with highly unevendepth. Bioinformatics 28, 1420–1428. doi: 10.1093/bioinformatics/bts174

Petersen, T. N., Brunak, S., von Heijne, G., and Nielsen, H. (2011). Signalp 4.0:discriminating signal peptides from transmembrane regions. Nat. Methods 8,785–786. doi: 10.1038/nmeth.1701

Petkau, A., Mabon, P., Sieffert, C., Knox, N. C., Cabral, J., Iskander,M., et al. (2017). Snvphyl: a single nucleotide variant phylogenomicspipeline for microbial genomic epidemiology. Microb. Genom. 3:e000116.doi: 10.1099/mgen.0.000116

Price, M. N., Dehal, P. S., and Arkin, A. P. (2009). Fasttree: computing largeminimum evolution trees with profiles instead of a distance matrix. Mol. Biol.Evol. 26, 1641–1650. doi: 10.1093/molbev/msp077

Quainoo, S., Coolen, J. P., van Hijum, S. A., Huynen, M. A., Melchers, W. J., vanSchaik, W., et al. (2017). Whole-genome sequencing of bacterial pathogens: thefuture of nosocomial outbreak analysis. Clin. Microbiol. Rev. 30, 1015–1063.doi: 10.1128/CMR.00016-17

Quick, J., Ashton, P., Calus, S., Chatt, C., Gossain, S., Hawker, J., et al. (2015). Rapiddraft sequencing and real-time nanopore sequencing in a hospital outbreak ofsalmonella. Genome Biol. 16:114. doi: 10.1186/s13059-015-0677-2


https://doi.org/10.1038/nbt.2522https://doi.org/10.1371/journal.pone.0104984https://doi.org/10.1128/AAC.02014-15https://doi.org/10.3389/fmicb.2017.00375https://doi.org/10.1093/bioinformatics/btr545https://doi.org/10.1101/gr.215087.116https://doi.org/10.1093/nar/gkm160https://doi.org/10.1016/B978-0-12-809633-8.20325-7https://doi.org/10.1186/1471-2105-11-461https://doi.org/10.1093/nar/gkh152https://doi.org/10.1099/ijsem.0.000760https://doi.org/10.1093/bioinformatics/btw152https://doi.org/10.1186/s40168-018-0480-xhttps://doi.org/10.1186/1471-2105-14-S5-S12https://doi.org/10.2903/sp.efsa.2018.EN-1498https://doi.org/10.1099/mgen.0.000135https://doi.org/10.1016/j.ygeno.2017.12.011https://doi.org/10.4081/idr.2017.6839https://doi.org/10.1093/nar/gkh435https://doi.org/10.1038/nmeth976https://doi.org/10.1186/1471-2105-9-386https://doi.org/10.2807/156https://doi.org/10.2807/1560-7917.ES.2017.22.23.30544https://doi.org/10.1128/JCM.01260-18https://doi.org/10.1101/530824https://doi.org/10.1093/nar/gkl723https://doi.org/10.1093/dnares/dsn027https://doi.org/10.1186/s13059-016-0997-xhttps://doi.org/10.1093/nar/gkt1226https://doi.org/10.1093/bioinformatics/btv421https://doi.org/10.1186/1471-2105-11-33https://doi.org/10.1371/journal.pone.0178154https://doi.org/10.1016/j.ijfoodmicro.2018.02.023https://doi.org/10.1093/bioinformatics/bts174https://doi.org/10.1038/nmeth.1701https://doi.org/10.1099/mgen.0.000116https://doi.org/10.1093/molbev/msp077https://doi.org/10.1128/CMR.00016-17https://doi.org/10.1186/s13059-015-0677-2https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles


Richter, M., Rosselló-Móra, R., Oliver Glöckner, F., and Peplies, J. (2016).Jspeciesws: a web server for prokaryotic species circumscriptionbased on pairwise genome comparison. Bioinformatics 32, 929–931.doi: 10.1093/bioinformatics/btv681

Roosaare, M., Vaher, M., Kaplinski, L., Möls, M., Andreson, R., Lepamets, M., et al.(2017). Strainseeker: fast identification of bacterial strains from raw sequencingreads using user-provided guide trees. PeerJ 5:e3353. doi: 10.7717/peerj.3353

Rosen, G., Garbarine, E., Caseiro, D., Polikar, R., and Sokhansanj, B. (2008).Metagenome fragment classification using n-mer frequency profiles. Adv.Bioinform. 2008:205969. doi: 10.1155/2008/205969

Saitou, N., and Nei, M. (1987). The neighbor-joining method: a new method forreconstructing phylogenetic trees.Mol. Biol. Evol. 4, 406–425.

Sarovich, D. S., and Price, E. P. (2014). Spandx: a genomics pipeline forcomparative analysis of large haploid whole genome re-sequencing datasets.BMC Res. Notes 7:618. doi: 10.1186/1756-0500-7-618

Schloss, P. D., Westcott, S. L., Ryabin, T., Hall, J. R., Hartmann, M.,Hollister, E. B., et al. (2009). Introducing mothur: open-source, platform-independent, community-supported software for describing and comparingmicrobial communities. Appl. Environ. Microbiol. 75, 7537–7541.doi: 10.1128/AEM.01541-09

Sedlar, K., Kupkova, K., and Provaznik, I. (2017). Bioinformatics strategiesfor taxonomy independent binning and visualization of sequencesin shotgun metagenomics. Comput. Struct. Biotech. J. 15, 48–55.doi: 10.1016/j.csbj.2016.11.005

Seemann, T. (2014). Prokka: rapid prokaryotic genome annotation. Bioinformatics30, 2068–2069. doi: 10.1093/bioinformatics/btu153

Segata, N., Waldron, L., Ballarini, A., Narasimhan, V., Jousson, O., andHuttenhower, C. (2012). Metagenomic microbial community profilingusing unique clade-specific marker genes. Nat. Methods 9, 811–814.doi: 10.1038/nmeth.2066

Sekse, C., Holst-Jensen, A., Dobrindt, U., Johannessen, G. S., Li,W., Spilsberg, B., etal. (2017). High throughput sequencing for detection of foodborne pathogens.Front. Microbiol. 8:2029. doi: 10.3389/fmicb.2017.02029

Sharma, A. K., Gupta, A., Kumar, S., Dhakan, D. B., and Sharma, V. K. (2015).Woods: a fast and accurate functional annotator and classifier of genomic andmetagenomic sequences. Genomics 106, 1–6. doi: 10.1016/j.ygeno.2015.04.001

Sharma, P., Satorius, A. E., Raff, M. R., Rivera, A., Newton, D. W., andYounger, J. G. (2014). Multilocus sequence typing for interpreting bloodisolates of staphylococcus epidermidis. Int. Perspect. Infect. Dis. 2014:787458.doi: 10.1155/2014/787458

Shimada, M. K., and Nishida, T. (2017). A modification of the phylip program:a solution for the redundant cluster problem, and an implementation of anautomatic bootstrapping on trees inferred from original data.Mol. Phylogenet.Evol. 109, 409–414. doi: 10.1016/j.ympev.2017.02.012

Silva, M., Machado, M. P., Silva, D. N., Rossi, M., Moran-Gilad, J., Santos, S., etal. (2018). chewbbaca: a complete suite for gene-by-gene schema creation andstrain identification.Microb. Genom. 4. doi: 10.1099/mgen.0.000166

Souvorov, A., Agarwala, R., and Lipman, D. J. (2018). Skesa: strategick-mer extension for scrupulous assemblies. Genome Biol. 19:153.doi: 10.1186/s13059-018-1540-z

Stamatakis, A., Ludwig, T., and Meier, H. (2005). Raxml-iii: a fastprogram for maximum likelihood-based inference of large phylogenetictrees. Bioinformatics 21, 456–463. doi: 10.1093/bioinformatics/bti191

Suvorov, A., Hochuli, J., and Schrider, D. (2019). Accurate inference of treetopologies from multiple sequence alignments using deep learning. bioRxiv.doi: 10.1101/559054

Tatusova, T., Ciufo, S., Federhen, S., Fedorov, B., McVeigh, R., O’Neill, K., et al.(2015). Update on refseq microbial genomes resources. Nucleic Acids Res. 43,D599–D605. doi: 10.1093/nar/gku1062

Tebani, A., Afonso, C., Marret, S., and Bekri, S. (2016). Omics-based strategies inprecision medicine: Toward a paradigm shift in inborn errors of metabolisminvestigations. Int. J. Mol. Sci. 17: E1555. doi: 10.3390/ijms17091555.

Treangen, T. J., Ondov, B. D., Koren, S., and Phillippy, A. M. (2014).The harvest suite for rapid core-genome alignment and visualizationof thousands of intraspecific microbial genomes. Genome Biol. 15:524.doi: 10.1186/PREACCEPT-2573980311437212

Vangay, P., Steingrimsson, J., Wiedmann, M., and Stasiewicz, M. J. (2014).Classification of listeria monocytogenes persistence in retail delicatessenenvironments using expert elicitation and machine learning. Risk Anal. 34,1830–1845. doi: 10.1111/risa.12218

Wheeler, N. E., Gardner, P. P., and Barquist, L. (2018). Machine learning identifiessignatures of host adaptation in the bacterial pathogen salmonella enterica.PLoS Genet. 14:e1007333. doi: 10.1371/journal.pgen.1007333

WHO (2014). Antimicrobial Resistance: Global Report on Surveillance. Technicalreport, WHO.

Wood, D. E., and Salzberg, S. L. (2014). Kraken: ultrafast metagenomicsequence classification using exact alignments. Genome Biol. 15:R46.doi: 10.1186/gb-2014-15-3-r46

Yuan, C., Lei, J., Cole, J., and Sun, Y. (2015). Reconstructing 16srrna genes in metagenomic data. Bioinformatics 31, i35–i43.doi: 10.1093/bioinformatics/btv231

Zankari, E., Hasman, H., Cosentino, S., Vestergaard, M., Rasmussen, S., Lund,O., et al. (2012). Identification of acquired antimicrobial resistance genes. J.Antimicrob. Chemother. 67, 2640–2644. doi: 10.1093/jac/dks261

Zerbino, D. R., and Birney, E. (2008). Velvet: algorithms for de novoshort read assembly using de bruijn graphs. Genome Res. 18, 821–829.doi: 10.1101/gr.074492.107

Zhang, S., Li, S., Gu, W., den Bakker, H., Boxrud, D., Taylor, A., et al. (2019).Zoonotic source attribution of salmonella enterica serotype typhimuriumusing genomic surveillance data, united states. Emerg. Infect. Dis. 25, 82–91.doi: 10.3201/eid2501.180835

Zhang, S.-W., Jin, X.-Y., and Zhang, T. (2017). Gene prediction inmetagenomic fragments with deep learning. BioMed Res. Int. 2017:4740354.doi: 10.1155/2017/4740354

Zhu, X., Leung, H. C. M., Chin, F. Y. L., Yiu, S. M., Quan, G., Liu, B.,et al. (2014). Perga: a paired-end read guided de novo assembler forextending contigs using svm and look ahead approach. PLoS ONE 9:e114253.doi: 10.1371/journal.pone.0114253

Conflict of Interest Statement: BV is the CEO of net-OMICS, a bioinformaticscompany.

The remaining authors declare that the research was conducted in the absence ofany commercial or financial relationships that could be construed as a potentialconflict of interest.

Copyright © 2019 Vilne, Meistere, Grantiņa-Ieviņa and Ķibilds. This is an open-

access article distributed under the terms of the Creative Commons Attribution

License (CC BY). The use, distribution or reproduction in other forums is permitted,

provided the original author(s) and the copyright owner(s) are credited and that the

original publication in this journal is cited, in accordance with accepted academic

practice. No use, distribution or reproduction is permitted which does not comply

with these terms.


https://doi.org/10.1093/bioinformatics/btv681https://doi.org/10.7717/peerj.3353https://doi.org/10.1155/2008/205969https://doi.org/10.1186/1756-0500-7-618https://doi.org/10.1128/AEM.01541-09https://doi.org/10.1016/j.csbj.2016.11.005https://doi.org/10.1093/bioinformatics/btu153https://doi.org/10.1038/nmeth.2066https://doi.org/10.3389/fmicb.2017.02029https://doi.org/10.1016/j.ygeno.2015.04.001https://doi.org/10.1155/2014/787458https://doi.org/10.1016/j.ympev.2017.02.012https://doi.org/10.1099/mgen.0.000166https://doi.org/10.1186/s13059-018-1540-zhttps://doi.org/10.1093/bioinformatics/bti191https://doi.org/10.1101/559054https://doi.org/10.1093/nar/gku1062https://doi.org/10.3390/ijms17091555.https://doi.org/10.1186/PREACCEPT-2573980311437212https://doi.org/10.1111/risa.12218https://doi.org/10.1371/journal.pgen.1007333https://doi.org/10.1186/gb-2014-15-3-r46https://doi.org/10.1093/bioinformatics/btv231https://doi.org/10.1093/jac/dks261https://doi.org/10.1101/gr.074492.107https://doi.org/10.3201/eid2501.180835https://doi.org/10.1155/2017/4740354https://doi.org/10.1371/journal.pone.0114253http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/http://creativecommons.org/licenses/by/4.0/https://www.frontiersin.org/journals/microbiologyhttps://www.frontiersin.orghttps://www.frontiersin.org/journals/microbiology#articles

Machine Learning Approaches for Epidemiological Investigations of Food-Borne Disease Outbreaks1. Introduction2. Machine Learning for de novo Microbial Genome Assembly3. Machine Learning for Microbial Genome Characterization3.1. Bacterial Strain Identification3.2. Bacterial Genome Annotation3.3. Virulence Gene Detection3.4. Antimicrobial Resistance Gene Detection

4. Machine Learning for Microbial Comparative Genomics4.1. Reference-Based SNP Methods4.2. Reference-Free SNP Analysis4.3. Pangenome-Based Analysis4.4. Core Genome/Whole-Genome Multi-locus Sequence Typing (MLST)

5. Machine Learning for the Inference of Microbial Phylogenomics6. ConclusionsAuthor ContributionsFundingReferences

Documents

Machine Learning Approaches for Epidemiological Investigations of Food-Borne Disease … · 2019. 9. 23. · Food-Borne Disease Outbreaks Baiba Vilne1,2*, Irena Meistere¯ 1, Lelde